As a public health researcher, I am primarily interested in the “so what?” of behavioural science. I want to know what solutions are available to solve the problems we so often identify in research. Thus, I found myself reading Pepper & Nettle's (P&N's) eloquent and convincingly argued target article, wondering what the dénouement would be in terms of implications for interventions.

The authors identify that reducing extrinsic sources of mortality to increase perceived control is likely to be “best achieved by addressing the social-structural inequalities” (sect. 7, para. 2). However, they do not linger long on this option, presumably considering wholesale social and political change to be beyond the reach of most behavioural scientists. Otherwise, they are concise and somewhat circumspect. One of their few concrete suggestions is to manipulate perceived control, but they acknowledge that it may be unethical to manipulate perceptions beyond the objective “truth.” If wholesale social and political change is off the table, then perhaps we must accept that persistent socioeconomic differences in objective control will persist. This circumstance leaves the potential for manipulating perceived control within the bounds of the “truth” somewhat limited.

However, maybe there is a middle ground. Perhaps it is possible to manipulate objective – and, hence, perceived – control without resorting to wholesale social and political change. Reflecting P&N's proposals, in the field of dietary public health, growing evidence suggests that many parents find it hard to enact their desire to feed their families and themselves better for reasons beyond their control. For example, mothers living in low-income neighbourhoods describe wanting to provide more fruits and vegetables for their children but not having the money to afford them or the time to prepare them (Jabs et al. Reference Jabs, Devine, Bisogni, Farrell, Jastran and Wethington2007). Qualitative interviews and preliminary outcome evaluations reveal that programmes that subsidise fruit and vegetable purchases can literally help mothers put fruit on the table (Bowling et al. Reference Bowling, Moretti, Ringelheim, Tran and Davison2016; Cohen et al. Reference Cohen, Richardson, Heisler, Sen, Murphy, Hesterman, Davis and Zick2017; Lloyd Reference Lloyd2014; McFadden et al. Reference McFadden, Green, Williams, McLeish, McCormick, Fox-Rushby and Renfrew2014). These findings suggest that increasing the financial availability of healthier foods increases parents' ability to enact the choices they want to make. In other words, and perhaps as P&N would rather put it, these interventions increase objective and perceived control.

Other interventions might also be expected to increase control over diets. These include changing the cognitive availability of healthier versus less-healthy food via restrictions on food marketing (Galbraith-Emami & Lobstein Reference Galbraith-Emami and Lobstein2013) and changing food's physical availability via planning restrictions on hot-food takeaways (National Institute for Health and Care Excellence 2013). Parents have identified ubiquitous exposure to advertising for less-healthy foods and hot-food takeaways as undermining their intentions to ensure their children eat well (Burges Watson et al. Reference Burges Watson, Tyrrell, Adamson, White and Adams2013; Khanom et al. Reference Khanom, Hill, Morgan, Rapport, Lyons and Brophy2015).

Adams et al. (Reference Adams, Mytton, White and Monsivais2016) have previously argued that what unites these interventions that support people to enact the choices they want to make is that they are “low agency”– that is, they do not rely on potential recipients' using their personal resources, or “agency,” to benefit. Instead, they act by changing the physical, financial, and cognitive availability of healthier choices.

In contrast, higher-agency interventions are generally focused on providing information via a variety of sources, including individual or group-based counselling, the mass media, social marketing campaigns, and labelling. Such information-based interventions rely on recipients' not just reading or listening to and remembering information but also being able to act on the changes that the information advocates. In environments where only some choices are physically, financially, or cognitively available, implementing this information can be nearly impossible. It is not difficult to see why these circumstances might undermine individuals' perceptions of the control they have over their diets. Indeed, it is difficult not to conclude that individuals in these circumstances truly have little control over their diets.

It is clear that in many contexts, fruits and vegetables are more expensive than other foods that are less healthy (Jones et al. Reference Jones, Conklin, Suhrcke and Monsivais2014; Monsivais et al. Reference Monsivais, Aggarwal and Drewnowski2012). When this is the case, only those with financial resources to spare (i.e., the less deprived) are able to implement advice to eat more fruits and vegetables. Similarly, when any aspect of the financial, physical, or cognitive environment makes information-based interventions harder for those living in more-deprived circumstances to implement, such interventions are likely to exacerbate existing socioeconomic inequalities in behaviour (Adams et al. Reference Adams, Mytton, White and Monsivais2016). This situation reflects P&N's proposal that “public health campaigns designed to elicit healthier behaviour by highlighting risks may actually decrease health effort if those risks are perceived to be beyond individual control” (sect. 7, para. 5). White et al. (Reference White, Adams, Heywood and Babones2009) have described these inequities in response to higher-agency interventions as “intervention-generated inequalities” (p. 68).

One of the common concerns with lower-agency public health interventions is that they impinge on individual choice (Frieden Reference Frieden2010; Nuffield Council on Bioethics 2007). Such interventions as soda taxes, regulation of food advertising, improvements to the nutritional quality of food served by takeaways, and changes in the density of such takeaways have all received the “nanny state” criticism from various quarters (Anon 2014; Henderson et al. Reference Henderson, Coveney, Ward and Taylor2009; Jou et al. Reference Jou, Niederdeppe, Barry and Gollust2014; Littlejohn Reference Littlejohn2008). And this criticism is not just journalistic rhetoric – the concerns about constraining choice and control that are wrapped up in the “nanny-state” label can decisively influence whether interventions are implemented (Giles et al. Reference Giles, Sniehotta, McColl and Adams2016). Importantly, P&N's thesis makes clear that by increasing recipients' control over their health behaviours, lower-agency interventions are likely to increase choice rather than decrease it.

From an applied public health point of view, the conclusion that lower-agency interventions can increase choice and control is perhaps the most important implication of P&N's article. The common, but flawed, argument that lower-agency public health interventions constrain choice and control needs to be clearly and consistently challenged.